Laser Raman spectroscopy-based characterization of metabolic heterogeneity in lymphoplasmacytic lymphoma and diffuse large B-cell lymphoma.

Raman spectral analysis is increasingly employed for tumor characterization; however, its application to the serum-based evaluation of non-Hodgkin lymphoma, particularly lymphoplasmacytic lymphoma/Waldenström macroglobulinemia (LPL/WM) and diffuse large B-cell lymphoma (DLBCL), remains relatively unexplored. To investigate a simple, non-invasive serum screening strategy for distinguishing LPL from DLBCL, we applied laser Raman spectroscopy in combination with orthogonal partial least squares discriminant analysis (OPLS-DA) to sera from the peripheral blood of seven patients with LPL, five patients with DLBCL, and five healthy controls. Classification models were subsequently established to differentiate lymphoma cases from healthy individuals and further to distinguish LPL from DLBCL. Compared with healthy controls, patients with lymphoma showed distinct serum Raman spectral patterns, characterized by decreased peak intensities at wavenumbers with contributions from nucleic acids (1485, 1491, 1573, 1579 cm-1), proteins (621, 643, 935, 1603, 1616, 1654, 1660 cm-1), and matrix-related proteins including collagen (859 cm-1). Increased intensities were observed in bands with contributions from lipids (1119, 1268, 1285, 1437, 1443, 1446 cm-1) and from carbohydrates and other biomolecules (842, 920, 1123 cm-1). Evaluation of routine serum metabolic parameters supported these spectroscopic findings, showing reduced levels of hemoglobin and high-density lipoproteins in lymphoma patients. In summary, these results identify characteristic spectroscopic signatures associated with LPL and DLBCL, and demonstrate the potential value of integrating Raman spectroscopy with conventional laboratory indicators for early, non-invasive disease assessment. This exploratory study suggests the utility of Raman spectroscopy as a complementary approach for the clinical evaluation and detection of LPL and DLBCL.